The Signal #4: Your RAID registers track four things. They miss the connections between them.
The Signal
You have Actions, Risks, Issues, and Decisions. If you’re a PM on any project that matters, you maintain some version of all four. Maybe it’s four tabs in a Google Sheet. Maybe it’s two spreadsheets and a section in your status report. Maybe it’s a Notion database, a Jira backlog, and a mental model you reconstruct every Monday morning.
The problem isn’t that you don’t track them. The problem is that you track them separately. And separately means independently. Which means the risk that materialised last Tuesday exists in your issues log, but the three actions it generated don’t link back to it. The decision that was made in Thursday’s steering committee has follow-up actions that don’t appear in your action register. The action that’s been overdue for two sprints is creating a risk that nobody has written down.
RAID was never designed as four separate documents. It’s one operating system. Risks create issues. Issues generate actions. Actions require decisions. Decisions introduce new risks. It’s a cycle - and every time you manage them in isolation, you break the chain that makes the framework useful.
The PM who maintains four registers separately is doing quadruple the work for half the insight.
The Noise
“We need separate owners for each register.”
Separate owners is fine. Separate systems is the problem. When your risk register doesn’t know about your action register, you’ve got four people maintaining four documents that don’t talk to each other - and a PM stitching it all together manually every week.
The Tool
Here’s the starter prompt. Paste in your meeting notes, Jira exports, Slack threads - whatever you have. One pass. Four registers.
You are a Senior RAID Analyst. Scan this project data and extract
every Action, Risk, Issue, and Decision - including ones buried in
casual language.
For each item, produce a row in the appropriate table:
ACTIONS:
Action ID | Description | Owner | Due Date | Priority | Status | Source
RISKS:
Risk ID | Description | Category | Probability | Impact | Severity |
Owner | Mitigation | Source
ISSUES:
Issue ID | Description | Severity | Blocker? | Owner | Source
DECISIONS:
Decision ID | Description | Decision Maker | Rationale | Follow-Up
Actions | Source
Rules:
- Scan for ALL four types in a single pass
- Flag cross-register links: "This action was generated by ISS-003"
- If no owner stated: [UNOWNED - PM to assign]
- Confidence: HIGH / MEDIUM / LOW for each item
PROJECT DATA:
[Paste your data here]You’ll get four tables. Scan them once. The cross-register links - “this action came from that issue, which materialised from that risk” - are the part you’ve never had before without manually drawing the lines yourself.
The Full Toolkit
That prompt is one of five. The complete free pack covers:
- RAID Extractor - all four registers from one data dump
- Action Tracker - every commitment with owner and due date
- Risk Classifier - detect and score risks using trigger patterns
- Issue Escalator - flag active problems and identify blockers
- Decision Logger - decisions with authority, rationale, and follow-up
>> Get all 5 free → projectorpm.xyz/raid-log-processor
P.S. For PMs already using the Meeting Minutes Manager: this product is where those meeting outputs go. Actions, risks, issues, and decisions all extracted from the meeting - now registered, scored, and linked as one system. The paid version adds a skill and workflow file that run the full RAID update cycle in Claude, ChatGPT, or Gemini - paste in your data, get all four registers updated in one pass.
The Signal is a no-filler newsletter by ProjectorPM. One practical AI insight for project managers, as often as there’s something worth saying. Unsubscribe any time.
Yes, AI helped me to write this :)